Importance of method validation: Implications of non-correlated observables in sweet taste receptor studies
Validation of research methodology is critical in research design. Correlation between experimental observables must be established before undertaking extensive experiments or propose mechanisms. This article shows that, observables in the popular calcium flux strength assay used in the characterization of sweetener-sweet taste receptor (STR) interaction are uncorrelated. In pursuit to find potential sweeteners and enhancers, calcium flux generated via G-protein coupling for wildtype and mutant receptors expressed on cell surface is measured to identify and localize sweetener binding sites. Results are channeled for sweetener development with direct impact on public health. We show that flux strength is independent of EC50 and sweet potency. Sweet potency-EC50 relation is non-linear and anti-correlated. Single point mutants affecting receptor efficiency, without significant shift in EC50 have been published, indicating flux strength is independent of ligand binding. G-protein coupling step is likely observed in the assay. Thus, years have been spent generating uncorrelated data. Data from uncorrelated observables does not give meaningful results. Still, majority of research in the field, uses change in calcium flux strength to study the receptor. Methodology away from flux strength monitor is required for sweetener development, reestablish binding localization of sweeteners established by flux strength method. This article serves to remind researchers to validate methodology before plunging into long term projects. Ignoring validation test on methodology, have been a costly mistake in the field. Concepts discussed here is applicable, whenever observable in biological systems are many steps moved from the event of interest.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Heckman-Corrected Epistemic Uncertainty: Selection on Unobservables Defeats Importance Weighting
Training data for machine learning is routinely collected by a selection process the model never sees: loans are observed only when granted, outcomes only when a test was ordered. The standard fixes -- importance weighti…
Applications of flow models to the generation of correlated lattice QCD ensembles
Machine-learned normalizing flows can be used in the context of lattice quantum field theory to generate statistically correlated ensembles of lattice gauge fields at different action parameters. This work demonstrates h…
Toward Supporting Perceptual Complementarity in Human-AI Collaboration via Reflection on Unobservables
In many real world contexts, successful human-AI collaboration requires humans to productively integrate complementary sources of information into AI-informed decisions. However, in practice human decision-makers often l…
Slower Generalization, Faster Memorization: A Sweet Spot in Algorithmic Learning
Critical-data-size accounts of grokking suggest a natural post-threshold intuition: once training data is sufficient to identify the underlying rule, additional data should accelerate validation convergence. We show that…
Doctor Rashomon and the UNIVERSE of Madness: Variable Importance with Unobserved Confounding and the Rashomon Effect
Variable importance (VI) methods are often used for hypothesis generation, feature selection, and scientific validation. In the standard VI pipeline, an analyst estimates VI for a single predictive model with only the ob…